[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Open Source Metadata Framework



Open Source Metadata Framework -OMF

Help test and refine the OMF templates by entering your documents and
other resources; go to http://metalab.unc.edu/osrt/omf/

What is the OMF metadata framework? Think card catalogue format that is:

completely open
standards-based (Dublin Core and XML)
interoperable (because of standards)
easily sharable (so you can put it in any database you choose)
author/contributor maintainable
includes XML validation 
includes highly flexible template entry
agnostic as to: 
	original manuscript format
	original manuscript content
	original manuscript location
	project in which the manuscript is located
	
OMF's metadata can be used to describe nearly any kind of resource:
SGML/XML, HTML, video, audio or plain text (again like the old card
catalogue). While the format is noted in the metadata, the metadata itself
is plaintext in XML. We want the metadata to be in XML for all the obvious
reasons. We built the template, description on Dublin Core for maximum
interoperability and sharing with other projects.

Even though we began by looking that the LDP when constructing the
metadata description, we think the description will work with all kinds of
Open Source Documentation projects. We have already been in conversation
with Deb Richardson of OWSG in order to make sure OMF interoperates with
the OSWG and the LDP. In fact, the structure of OSWG's OSDI are based on
the OMF schema so that entries in the OSDI could easily be extracted as
XML based OMF records and shared through OMF.


We see no reason why OMF would preempt any full text project such as the
LKB or even ODE. Recognizing that all documentation might not make it into
LKB or even ODE, we hope that the metadata will be the pathfinder to all
kinds of Open Source resources.

We plan to host at least one search/retrieval application based on OMF,
but we see the metadata as being open and shared so that alternative
schemes, homes, searching plans, and value-added, like OSDI'9s reader
comments, are welcome and encouraged.

In many ways, the OMF is somewhat analogous to the very successful LSM
(Linux Software Maps) which we used for the Metalab Linux Archives since
the early 90s. Searching then is analogous to Linsearch,
http://metalab.unc.edu/linsearch/ which allows you to do a template-based
search for software no matter where it is based, what language it was
written in, etc. Of course, there are several other LSM search stategies
implemented at other sites. Since we all share the LSMs, people can choose
whatever searching scheme best suits them.


To discuss the OMF and other metadata issues with the developers and to
participate in future development, subscribe to our mailing list by
sending a message of:  subscribe ldp-meta to: franklin.oit.unc.edu

Help test and refine the template by entering your documents and other
resources; go to http://metalab.unc.edu/osrt/omf/

All OMF tools will be availabe via tarball, thanks to Kendall Clark, in
the near future.

OMF Team:
Kendall Clark <[email protected]>
Miles Efron <[email protected]>
Jane Greenberg <[email protected]>
Paul Jones <[email protected]>

With help and advice from:
Jesus M. Gonzalez-Barahona <[email protected]>
Matt Knuppel <[email protected]>
Jonathan Magid <[email protected]>
Ismael Olea <[email protected]>
Victor R. Ruiz <[email protected]>


--  
To UNSUBSCRIBE, email to [email protected]
with a subject of "unsubscribe". Trouble? Contact [email protected]